Virtual presenter

  • Authors:
  • Denise Hippler;José Mario De Martino

  • Affiliations:
  • State University of Campinas, Campinas, SP, Brazil;State University of Campinas, Campinas, SP, Brazil

  • Venue:
  • Proceedings of the VIII Brazilian Symposium on Human Factors in Computing Systems
  • Year:
  • 2008

Quantified Score

Hi-index 0.00

Visualization

Abstract

Virtual presenters have a great range of possible applications, like teachers, news presenters and guides of virtual environments, easing the interaction with computers. The animation control of such virtual characters is usually accomplished by an animation script describing all the movements to be performed. Writing a convincing animation script is a demanding and cumbersome task. To ease the animation process, we propose the additional use of a behavior model learned from a real presenter. The article presents the implementation of a 3D virtual news presenter that implicitly follows a behavior model and a script that describes the text to be uttered. The behavior model is given by a set of behavioral rules that represent common non-verbal facial movement patterns displayed by a real presenter whose TV appearances were analyzed